How your data stays protected using RoseRx

How RoseRx handles your data

When RoseRx answers a question from an HCP or a patient, it sends the relevant approved content and the question itself to an AI model, and the model sends back a response. We set this up so that the data sent to the model is never stored afterward and never used to train that model. This protection lives with RoseRx, not with whichever AI provider we’re using, so it doesn’t depend on their policies and it doesn’t change if we switch to a different model.

Why this matters

AI models are built by training on huge amounts of data, and once that happens, the model can sometimes repeat parts of what it learned. So if an AI company trained its model on the data one customer sent it, that data could show up in an answer given to a completely different customer. Most large AI providers already say they won’t do this for business customers, but whether that actually holds depends on how the connection is set up and kept that way over time.

For a pharma company, the data going into these conversations may include sensitive data like patient information, adverse event details, or the content of how a brand talks to doctors and patients. Some of this carries legal obligations, and some of it is commercially valuable because it shows what is working. None of it should end up stored by an AI provider or visible in someone else’s results.

The data belongs to you

RoseRx stores and processes it on your behalf and under your control, in our own systems rather than scattered across the AI providers we use. You keep ownership of the content, the patient information, and the record of what was asked and answered. We hold it so that it can be governed in one place, kept out of model training, and produced for your privacy and security teams whenever they need to see how it is handled.

Switching models without switching risk

Because we handle data protection ourselves rather than relying on the AI provider, we can change which model we use without changing how your data is handled. AI models are improving and changing price constantly, and the best one at any given time might come from a provider your legal or security team wouldn’t approve for direct use, sometimes because that provider operates under another country’s data laws. Since we already keep your data protected at our end, we can move to a better model as one comes along without sending your data to that provider directly and without you having to redo a security review every time.

How we help your team

RoseRx is audited and certified to ISO 42001, the international standard for managing AI systems. We can walk your privacy, security, or governance team through exactly how this works.